UNIVERSITY OF THE WITWATERSRAND

Transcription

1 AETIOLOGY OF CONJUNCTIVITIS IN SEBOKENG HOSPITAL NEONATES WHO RECEIVED ROUTINE PROPHYLAXIS AT BIRTH AGAINST EYE INFECTION DR CHERIAN THOMAS STUDENT NO : G RESEARCH REPORT SUBMITTED IN PART FULFILLMENT OF MSc (Med) IN CHILD HEALTH UNIVERSITY OF THE WITWATERSRAND I

2 DECLARATION I, D r Cherian Thomas, hereby declare that this research report is my own work. It has not been submitted to any other University before for any degree or examination. It is being submitted for the degree of Master of Science in Medicine (Child Health), University of the Witwatersrand, Johannesburg. D R C. THOMAS DATE : 8 06'I 99 II

3 DEDICATION I dedicate this research report to my late parents, Dr M. Thomas and Dr (Ms) Alice Thomas, my dear wife Anita, my children Thomas Junior and Serra. HI

4 ABSTRACT Introduction Neonatal conjunctivitis remains an important cause of morbidity, particularly in developing countries, despite the widespread use of antibacterial prophylaxis. Objectives 1. To establish the aetiology o f conjunctivitis among Sebokeng neonates. 2. To assess the efficacy o f the prophylaxis used at Sebokeng Hospital. 3. To propose an effective prophylaxis and treatment protocol for neonatal ophthalmia based on the study results. M ethod A prospective, observational study was carried out at Sebokeng Hospital, Gauteng. Hospital-born neonates presenting with conjunctivitis, between 0-30 days, who had routine prophylaxis (10% Spersamide) at birth were studied.. In addition, 10 home bom infants who did not receive prophylaxis and presented with neonatal conjunctivitis were also studied. The study was conducted over a 8 month period (November July 1996) Two eye swabs were collected from each neonate, one for routine culture and susceptibility tests and the other for chlamydial culture. Results were statistically analysed using Chi-squared contingency table tests. Mothers of affected infants were interviewed about the course o f their pregnancy and about possible predisposing factors for their infant s illness. IV

5 Results In the 105 hospital bora infants with neonatal conjunctivitis, the commonest pathogenic organisms isolated were Staphylococcus aureus in 23.8%, Chlamydia trachomatis in 12.4%, Neisr eria gonorrhoeae in 11.4% and Streptococcus pneumoniae in 9.5%. Gram negative bacilli were found in a total o f 22 (20.9%) patients, the commonest being Klebsiella pneumoniae The incidence of ophthalmia neonatorum was 3.66 cases per 100 live births; that of chlamydia 0.45 cases per 100 live births and the gonorrhoeae rate was 0.41 cases per 100 live births. Though Spersamide (10% sulphacetamide) eye drops appeared to function adequately as a prophylactic agent, its efficacy in treating N, gonorrhoeae. S, aureus and C trachomatis were inadequate. An overall 96% sensitivity rate to Chloramphenicol was shown by the commonest pathogenic organisms. Home-born babies who did not receive any prophylaxis showed a higher rate o f gonococcal and S. aureus infection rate (30% each) than hospital bora babies. However, this was not statistically significant owing to the small number o f neonates in this subset (10 neonates). Only 25 (20%) mothers were married. A history of vaginal discharge was present in 78% o f mothers, 68% of whom received some form of treatment antenatally. Conclusions S. aureus emerged as the most common cause o f eye infection among Sebokeng Hospital bom neonates followed by C. trachomatis and N. gonorrhoeae. Spersamide 10% eye drops was found to be inadequate as a prophylatic agent against eye infection. It is recommended that it should be replaced immediately by chloramphenicol eye drops or eye ointment. Recommendations are also provided on the optimal management o f established neonatal conjunctivitis. V

6 ACKNOWLEDGEMENTS I remember with gratitude all those who helped me at various stages of my research. A special mention needs to be made o f the following people. 1. My supervisor. P rof H. Crewe-Brown, who despite her busy schedule o f work always found time for me. 2. D r Anne van der Spuy (ex Superintendent), of Sebokeng Hospital and D r F.S. Kalla (Senior Consultant in Paediatrics) of Sebokeng Hospital, for their support during my course and research. 3. Medical Microbiologist M r J. Kruger, Ms P.P. Allen, Ms N. Shorkend and the entire staff o f SAIMR Laboratory o f Sebokeng Hospital for their co-operation. 4. Ms R. Nagel who helped me in typing and re-typing all the proofs and the final report. 5. The entire paediatric nursing staff. Primary Health Care nurses and Medical Officers o f Department o f Paediatrics o f Sebokeng Hospital for their cooperation. 6. Dr (Ms) J. S. Galpin, Department o f Statistics, University of Witwatersrand for her help. 7. My family who helped me during the whole length of my course in every way possible. 8. The illustration units of medical school and the SAIMR for assistance with graphic work. 9. Mr S.H. Kassim, Principal Roshnee School for helping me with the scientific editing o f the research report. I thank you all for your sincere help. D r C. Thomas VI

7 TABLE OF CONTENTS PAGE TITLE PAGE DECLARATION i ii LIST OF FIGURES LIST OF TABLES x xi 1.0 INTRODUCTION Causes o f neonatal conjunctivith Diagnosis Prophylaxis Treatment THE RESEARCH PROBLEM Objective METHODS Selection o f patients Collection and storage o f specimens Ethijal clearance Pilot study Statistical analyses RESULTS Organisms isolated Time o f presentation Antibiotic susceptability incidence Management of conjunctivit is Questionnaire results 38 VII

8 5.0 DISCUSSION Incidence Distribution of organisms Susceptability to drugs Treatment Questionnaire Recommendations Conclusion 66 REFERENCES 67 APPENDIX 75 VIII

9 LIST OF FIGURES FIGURE PAGE 1 Day o f presentation of neonatal conjunctivitis after birth - 32 hospital bom neonates 2 Week o f presentation o f conjunctivitis in hospital born neonates 33 in relation to important organisms isolated 3. Flow chart-guideline for treating neonatal conjunctivitis at Sebokeng hospital IX

10 LIST OF TABLES TABLES PAGE 1.1 Ophthalmia neonatorum : Table showing studies published 8 around the world comparing the efficacy o f various prophylactic agents 4.1 Bacteria isolated on culture o f swabs from hospital bom babies Summary o f organisms isolated from eye swabs o f hospital 29 bom neonates 4.3 Sensitivity results o f five major organisms isolated in the 34 study to various antibiotics 4.4 Bacteria isolated on culture o f swabs from home bom babies Organisms isolated from neonates non-responsive to first line 37 therapy 5.1 Relative prevalence o f fr aureus in different parts o f the world Comparison between Sebokeng and Zaragoza study 42

11 5.3 Incidence of chlamydia and gonococcus in various parts of the world 5.4 Relative prevalence of chlamydia from various studies Gonococcal conjunctivitis Relative prevalence in various 47 studies 5.6 Various prophalactic agents used and the controversy about their efficacy XI

12 INTRODUCTION Neonatal conjunctivitis is defined as inflammation of the conjunctiva in an infant younger than 1 month old. It is characterised by redness and swelling of the eyelids and palpebral conjunctiva and a purulent discharge, with one or more polymorphonuclear cells per high power field in a gram-stained conjunctival smear [I]. In 1881, prophylactic treatment for neonatal conjunctivitis was introduced by Crede. He used silver nitrate to prevent infection with Neisseria gonorrhoeae. [1] Over the past century, this practice has been maintained to prevent neonatal conjunctivitis. Although chemical and viral agents produce ophthalmia neonatorum, bacteria including chlamydia continue to play a major role in causing neonatal conjunctivitis [2-6]. The known causes o f neonatal conjunctivitis have continued to expand and now includes Staphylococcus. Streptococcus. Neisseria species, herpes virus as well as chemical conjuctivitis. This expansion has brought to light a few organisms which were not responsive to silver nitrate and this has prompted a search for new and broader prophylactic agents, e.g. povidone iodine, tetracycline eye ointment, erythromycin eye ointment, etc. 72

13 It has been reported that 42% of neonates exposed to N. gonorrhoeae during delivery and 31% exposed to Chlamydia trachomatis may develop conjunctivitis. Neonatal infection caused by herpes simplex virus develops in 40-50% of newborns exposed to active genital herpes virus [2], 1.1 CAUSES OF NEONATAL CONJUNCTIVITIS Chemical conjunctivitis Chemical conjunctivitis used to be the most common form of neonatal conjunctivitis, related to the prophylactic use o f 1% silver nitrate. It is self- limiting and resolves in 24 to 48 hours [1-2] Bacterial conjunctivitis As noted earlier, various bacteria continue to cause neonatal conjunctivitis despite prophylaxis. Traditionally N. gonorrhoeae has been blamed as the cause o f typical ophthalmia, neonatorum with varying grades of clinical presentation [2-6]. Despite significant improvements in the prevention o f conjunctivitis since antibacterial prophylaxis was introduced, N. gonorrhoeae still accounts for a significant number o f cases. It presents as a hyperacute purulent conjunctivitis with a variable onset [2], Other bacterial organisms can produce neonatal conjunctivitis because several types of bacteria can be found in the endocervix during early stages of labour [1, 7-9]. Gram positive organisms include Staphylococcus aureus. S. pneumoniae and groups A and B streptococci, gram negative /3

14 3 Escherichia coli. Klebsiella pneumoniae. Serratia marcescens as well as Proteus and Enterobacter species. The bacteria found in the conjunctiva of the newborn reflect the mode o f delivery (whether vaginal or by caesarian section), In infants delivered vaginally, the bacteria are characteristic o f the flora of the female genital tract [5, 10], In babies bom by caesarian section the conjunctival flora depends on the time elapsed between the membrane rupture and the operation [1], In a study of 100 newborns, examining the bacteria isolated from conjunctival cultures taken within 15 minutes of delivery, anaerobic organisms accounted for 78% of the positive cultures and aerobic organisms for 22%. [1] Frequently isolated anaerobes were Lactobacillus (41%) and Bifidobacterium species (19%) diphtheroids (6%), Propionibacterium acnes (6%) and Bacteroides species (5%). [2, 6l O f the aerobic organisms isolated, Staphylococcus ev idermidis are the most common (9%) followed by Corvnebacterium (6%), Streptococcus species (3%) and Escherichia coli 12%). [1] A study conducted in 1984 in a Nairobi hospital where no prophylaxis was used gives useful information on the likely organisms causing neonatal conjunctivitis in an African setting. Twenty-three percent of the neonates developed ophthalmia neonatorum. Gonococci accounted for 12% of cases, chlamydia for almost a third (32%), and a combination o f the two for 3%. In those cases of ophthalmia not caused by these two organisms, the most frequent organisms were Haemophilus influenzae, Staphyloccoccus aureus and Streptococcus pneumonia, each o f which caused 6-7% of cases. [11] /4

15 Viral conjunctivitis Neonatal herpes sim- iex virus infection may affect the eyes as we'l as the skin with oral and vaginal lesions. If the mother has a history o f a genital herpetic infection, neonatal herpes should be suspected. [1] In one reported series of 297 neonates with herpes simplex virus infection, ocular pathology was seen in 17% and conjunctivitis, specifically, in 10%. [1] 1.2 PATHOGENESIS AND PATHOGENETIC ROLE OF ORGANISMS CAUSING OPHTHALMIA NEONATORUM The incidence and pattern of colonization depend on patient age, geographic location and ambient climatic conditions. Although some commensal organisms may produce an opportunistic infection, the pathogenic species are more likely to cause clinical disease, particularly conjunctival infection [2], The clinical disease and the severity of symptoms, signs and complications depend on the pathogenesis of the disease process. This is determined by the growth characteristics o f the organism, the production o f toxins or enzymes by it and the type and degree of inflammatory response mounted by the host [2, 3], Bacteria usually need to adhere to the epithelial surface before they can initiate clinical disease. The intact epithelium is an effective barrier to most organisms with the exception o f N. gonorrhoeae. Corvnebacterium diphtheriae. Listeria and Haemophilus aegyptius. that can, through specialized surface attachments, penetrate an intact epithelium [2, 4],

16 - 5 - The specialized surface attachments of organisms like the pseudomonas and gonococci enable biologic adhesion to occur, to the glycocalyx o f injured epithelium, which then allows entry of diffusable toxin or bacterial products into the stroma [2]. A variety o f enzymes can be produced by bacteria (e.g. proteases, coagulases, nucleases, elastases, lipases and fibrinolysins, etc). They also produce various toxins e.g. oc, p and y toxin o f Staphylococcus. Disruption of the underlying tissue by the digestive effect o f the released enzymes then allows further access of the organisms into tissue where they protect themselves from host defenses [2,4], In the conjunctiva the pathogenetic options produce a broad spectrum o f diseases from self-limiting infections, such as a modest symptomatic conjunctival injection and chemosis, to severe inflammatory membranes that slough causing pain and photophobia. 1.3 DIAGNOSIS Because o f the potentially disastrous outcome of untreated neonatal conjunctivitis, the diagnosis o f the aetiological agent and appropriate treatment must be initiated as soon as possible [1], Though clinical signs and symptoms give minimal clues about the organism responsible for the infection, the time of onset may give some clues, e.g. conjunctivitis due to N. gonorrhoeae usually has its onset during the first two to /6

17 - 6 - five days o f life. Chlamydia trachomatis conjunctivitis usually has its onset during the second week. Signs and symptoms of conjunctivitis are obvious and vary in intensity with the organism responsible. In the case o f N. gonorrhoeae. within a day eyelids become oedematous, with prominent chemosis and the emission o f thick purulent discharge. In cases o f C. trachomatis, and other bacteria, symptoms and signs are highly variable, ranging from mild to severe oedema o f the eyelids with minimal or copious purulent discharge [7], Viral conjunctivitis due to herpes simplex virus may present within two weeks after birth and may be preceded by vesicular skin rashes, "t may be unilateral or bilateral. Presence o f microdendritic or geographical ulcers in the cornea are the hallmark of neonatal herpes conjunctivitis [1], However, because o f the variability in clinical presentation, symptoms and signs are unhelpful in making an aetiological diagnosis. Thus, microbiological culture and susceptability remains imperative for the identification o f the causative organism as well as establishing the most effective therapy. Culture specimens o f the conjunctival surface should be obtained without the use o f topical anaesthetic using a calcium alginate or dacron swab inoculated onto blood agar, chocolate agar and thioglycollate broth. If gonococcal infection is suspected it is wise to inoculate it onto a Thayer - Martin agar. [2] 77

18 - 7 - Conjunctival smears can then be obtained after using topical proparacaine anaesthesia by gently scraping the everted palpebral conjunctiva with a platinum spatula. Scraping can then be evaluated using Gram s and Giemsa stains. A smear from an infant with chemical conjunctivitis would be characterized by the presence o f neutrophils and occasional lymphocytes: whereas bacterial conjunctivitis would contain bacteria and neutrophils. Intracellular Gram negative diplococci are identifiable i'n up to 95% of Gram stains in eye culture positive cases o f gonococcal conjunctivitis [2], The smear from chlamydial conjunctivitis would show neutrophils, lymphocytes, plasma cells and basophilic intracytoplasmic inclusions in epithelial cells [10], In the case o f a viral conjunctivitis the smear would show lymphocytes, plasma cells, multi-nucleated cells and eosinophilic intranuclear inclusions [1, 12-14], Thus, prior to the culture results being available, a preliminary diagnosis can be made from the examination o f a smear. Viral cultures are only taken if the smears suggest it, and if there is evidence o f herpes infection in the mother, since they are expensive. 1.4 PROPHYLAXIS The use of 1% o f silver nitrate instilled in the conjunctiva shortly after birth has been the mainstay o f prophylaxis for neonatal conjunctivitis over the past century.,/8

19 - 8 - Concerns about the emergence o f organisms resistant to silver nitrate and the chemical conjunctivitis caused by it are reasons to look for alternatives [1, 12-13, 15]. Various studies have been carried out in different parts of the world using 0.5% erythromycin ophthalmic ointments, a 1% tetracycline ophthalmic ointment and povidone iodine (2.5% solution). The reports indicate that they are all as effective, or better than, silver nitrate as they do not cause chemical conjunctivitis and are as cost effective [1, 12-13], Nevertheless none of them are 100% effective even when administered appropriately. See table 1,1. TABLE 1.1 : OPHTHALMIA NEONATORUM : TABLE SHOWING STUDIES PUBLISHED AROUND THE WORLD COMPARING THE EFFICACY OF VARIOUS PROPHYLACTIC AGENTS COUNTRY. YEAR SUBJECT NO ORGANISM SILVER NITRATE +% TETRA CYCLINE % ERYTHRO MYCIN % NIL. % POVIDONE DDINE % USA (16) All organisms Not effective Not effective ZAIRE (17) All organisms KENYA (18) All organisms USA (19) Chlamvdia 33 0 USA (20) Chlamvdia TAIWAN ri3) Chlamvdia I 1.6. * Percentage represents effectiveness o f the drug to organisms cultured from eye swabs. 79

20 - 9 - In the prevention o f neonatal ocular infection, four levels o f intervention can be used [11,21-22], The first involves the prevention o f sexually transmitted diseases. The emergence o f the human immunodeficiency virus has led to new health promoting strategies. These programmes may reduce the prevalence of sexually transmitted diseases, which in turn could reduce the risk to infants o f exposure to the agents that cause neonatal conjunctivitis. The second approach consists of antenatal screening. Ophthalmia neonatorum can be prevented by screening pregnant women for genital infection, particularly those at high risk for disease. Foul-smelling vaginal discharges must be treated aggressively, following culture o f the organisms. [12]. The third approach is ocular prophylaxis at birth, which is simple and inexpensive. It consists o f cleaning an infant s eyelids with a dry swab as soon as possible after birth and then instilling a safe, available and affordable antimicrobial agent. Tetracycline eye ointment 1%, povidone iodine solution 2.5%, or chloramphenicol eyedrops are all equally effective as silver nitrate and cause less side effects (e.g. chemical conjunctivitis) [6, 9, 12, 15]. 710

21 Finally, early diagnosis and adequate treatment after identification o f the causative organism can prevent corneal ulceration and blindness. Elimination of childhood blindness due to ophthalmia neonatorum needs an interdisciplinary approach involving gynaecologists, neonatologists, ophthalmologists and most importantly primary care workers. All primary health care workers should be educated about the cause, prevention and treatment o f neonatal conjunctivitis. 1.5 TREATMENT Bacterial Conjunctivitis The standard therapy recommended for gonococcal neonatal conjunctivitis consisted of frequent topical aqueous penicillin and also systemically administered penicillin. With the emergence o f penicillin resistant gonococcal strains, this treatment has fallen out o f favour [1, 4, 12, 23]. Ceftriaxone in a single-dose intramuscular injection at 50mg/kg per 24 hours works extremely well, is quick and cost-effective. This also treats extra ocular gonococcal infection [24]. Co-infection with C. trachomatis is not unusual in gonococcal conjunctivitis. Treatment of chlamydial conjunctivitis must be aimed not only at ocular colonization but also toward eradication o f the nasopharyngeal carriage [3, 5, 15]. Oral erythromycin estolate or ethyl succinate suspension, mg/kg/day for days is the treatment of choice [7-8, 10, 25]. /II i

22 Viral Conjunctivitis Although the efficacy o f treatment o f herpes simplex conjunctivitis and blepharitis has not been fully established [1], the use o f topical trifluorothymidine should be considered. If there is no response, other anti-viral agents like acyclovir should be initiated locally and parenterally. 1.6 COMPLICATIONS Complications o f untreated neonatal conjunctivitis caused by N. gonorrhoeae were common prior to the institution o f ocular prophylaxis in the late nineteenth century. Until then neonatal conjunctivitis was the leading cause of blindness in children [1], Usual complications include comeal ulceration, perforation, iridocyclitis, anterior synechiae and panophthalmitis. Long term complications o f chlamydial conjunctivitis are rare, but this should be considered as a presenting sign o f chlamydial infection in the newborn, which m ay cause chlamydial pneumoniae, recta! and vaginal infection, etc. 712

23 THE RESEARCH PROBLEM It has been an established routine at Sebokeng Hospital to instill Spersamide 10% eye drops (a combination of sodium sulphacetamide and mercuric borate), into the eyes of all the hospital born neonates (3-5 drops in each eye). Despite this, many neonates continue to present at the hospital with conjunctivitis during their initial stay or after discharge. This results in the re-admission o f babies and their mothers. A belief prevalent among the doctors in the casualty department, is that all cases of neonatal conjunctivitis are caused by gonococci. This results in the empirical use o f intravenous crystalline penicillin and penicillin eye drops for all cases o f neonatal conjunctivitis. In many instances, this treatment regime has proved to be ineffective, inconvenient for the baby and the mother and has resulted in an unnecessarily prolonged stay in the hospital, which ultimately leads to a strain on the hospital budget. Treatment is often commenced without any eye swabs being taken for culture. This has resulted in poor understanding about the aetiology of conjunctivitis among these neonates and has impeded the establishment of an effective ocular prophylaxis and treatment protocol. The aim of this study was to establish the organisms causing neonatal conjunctivitis in the Sebokeng area. It would also evaluate the efficacy of Spersamide 10% in preventing ophthalmia neonatorum. This would also assist in establishing a more effective treatment protocol. /13

24 OBJECTIVES Primary objectives 1. To establish the aetiology o f conjunctivitis in neonates bom at Sebokeng Hospital. 2. To assess the efficacy o f the prophylaxis used at Sebokeng Hospital (Spersamide 10%) Secondary objectives 1. To estimate the incidence o f neonatal conjunctivitis in the Sebokeng area 2. To devise and propose a more cost-effective treatment protocol. 714

25 METHODS 3.1 DESIGN This was a prospective, descriptive, observational study. 3.2 DEFINITIONS Neonate: A patient between 0 to 30 days of age. Conjunctivitis: Redness and swelling of the eyelids and palpebral conjunctiva, sticky eyelids and an associated purulent discharge. 3.3 SELECTION OF PATIENTS Inclusion criteria Any neonates who developed a conjunctivitis during their stay in the hospital (after birth) Any neonate who presented with conjunctivitis at the paediatric out-patient department at Sebokeng hospital between OShOO and 16h00 from Monday to Thursday. Infants born outside the hospital but in the Sebokeng area, e.g. at home, were eligible for study, but were analysed separately (as they did not receive ocular prophylaxis) Exclusion criteria e Neonates with conjunctivitis who presented at the out-patient department after 16h00 each day from Friday to Sunday or on public holidays Parent unwilling to complete patient questionnaire. /IS

26 SAMPLE SIZE It was decided to enroll all patients presenting over a 8-month period (November July 1996). It was estimated that this would include 100 patients. COLLECTION AND STORAGE OF SPECIMENS 1 Specimen collection At the initial visit, two eye swabs were collected, using sterile swabs, from each infected eye. The first swab was sent for microscopy and bacterial culture and susceptibility testing to the local, hospital laboratory. The second swab was placed in a chlamydia transport medium, packed in ice, and sent to a central laboratory in Johannesburg (South African Institute for Medical Research [SAJMR]) for chlamydial culture. It was decided to confine the collection o f swabs to office hours from Mondays to Thursdays since the laboratory facility at Sebokeng Hospital was only available from 08h00 to 17h00 each day and not at all during weekends and public holidays. In addition, the chlamydial eye swabs had to be transferred to the central SAIMR laboratory which was 78 km av/ay. Transport of specimens between this facility and the hospital were recognised to be poor on weekdays and non-existent on weekends. The researcher collected all specimens and took them immediately to the local laboratory where a laboratory technician performed routine microscopy, culture and susceptibility tests. Slides for microscopy were prepared from the swabs and were stained with Gram s

27 Specimen processing The following media were used for inoculation: 1. Blood Agar 5% For streptococci, staphylococci, pneumococci 2. Chocolate Agar For Haemophilus spp and N. gonorrhoeae 3. Thayer Martin Agar For N. gonorrhoeae 4. Serum Broth For sub-culture for fastidious organisms 5. MaConkey Agar For Gram negative bacilli, e.g. Pseudomonas Organisms cultured on the media were identified by standard methods. For antibiotic susceptibility studies, disc susceptibility tests were used. In cases where N. gonorrhoeae and H. influenzae were isolated Beta-lactamase strips were used. For C. trachomatis culture, a cell culture technique (McCoy cells) was used by the SAIMR laboratory [26] Questionnaire Each infant s mother was interviewed. The questionnaire (a copy o f which is attached) recorded information on the baby s place of birth, clinical history, including the receipt o f ocular prophylaxis, and the mother s antenatal and perinatal history and its management.,/17

28 Collection o f incidence data To allow an estimate o f the incidence of neonatal conjunctivitis in the Sebokeng area, a record was kept o f all neonates who presented to the paediatric polyclinic, outside the study enrollment hours. All mothers, on discharge, from the hospital were also requested to come directly to the hospital if their child t "eloped signs of conjunctivitis. A printed notice, authorised by the Superintendent, was sent to the peripheral clinics to be displayed on the notice boards o f these clinics. The content of this letter was mainly an appeal to the Primary Health nurses to refer cases of neonatal conjunctivitis to Sebokeng Hospital s paediatric Polyclinic, without commencing any treatment. 5.5 Spersarnide 10% Eves drops The agent used for ocular prophylaxis in this study was Spersarnide 10%. Each 100 ml o f this substance contains sodium sulphacetamide log, hydroxypropylmethylcellulose 0.2g, phenylmercuric borate 0.001%, and sterile water to 100ml. 6 ETHICAL CLEARANCE Permission to do the study was obtained from the Senior Superintendent of the hospital and Head o f the Department o f Paediatrics at Sebokeng Hospital. The study protocol was submitted on to the University of the Witwatersrand Committee for Research in Human Subjects. It was approved unconditionally on the /18

29 The protocol was approved by the Postgraduate Committee on I4U Nov Copies of both approval letters are attached herewith. 3.7 PILOT STUDY A pilot study of 15 cases o f neonatal conjunctivitis was carried out from the 24th November 1995 to 2nd January The study was done according to the protocol steps explained previously. It was encouraging to note that using the study design, two cases o f positive chlamydial culture and several positive cultures, including Staphylococcus aureus. Staphylococcus enidermidis. Haemophilus influenzae and Neisseria gonorrhoeae were obtained. Relevant information was also collected using the questionnaire. A problem in transportation of the specimens was identified during the pilot study. It was noted from a few o f the laboratory reports that the central SAIMR laboratory were not always culturing chlamydia owing to the fact that some specimens reached them 24 hours after collection and that many specimens had also not been transported on ice. This failure occurred because the hospital transport system only delivered the specimens on two days (Monday and Thursday) to the central SAIMR. It was also found that collecting specimens on Fridays was futile since there was no one to deliver them the next day (Saturdays). In response to this Sun Couriers were contacted who agreed to transport specimens from Vereeniging Hospital f Johannesburg on weekdays. 719

30 This meant that the researcher had to deliver the specimens to Vereeniging Hospital each day. Moreover, all specimens were placed in plastic bags containing ice cubes so that when they reached SAIMR they were suitable for culture. 3.8 ALLOCATION OF TASKS A formal request for co-operation (referral o f patients) from colleagues and nurses in and around the hospital was made in writing. All other tasks were done by the researcher. This included the examination of neonates, interviewing of the mother, filling in o f the questionnaire, the collection o f eye swabs, ensuring the transportation of specimens to the laboratory and the collection o f results. The medical technologist at the hospital SAIMR laboratory offered his support in ensuring the prompt processing o f the specimens. 3.9 STATISTICAL ANALYSIS Most data were analysed manually. D r J.B. Galpin from the Department of Statistics at the University o f Witwatersrand helped with the statistical analysis. Chi-square contingency table tests were used when appropriate BUDGET There were no additional costs to the hospital as the study methodology formed part o f the routine management o f neonatal conjunctivitis. All other costs (eg. printing costs) were paid for by the researcher. No financial assistance was obtained from outside sources.

31 4.0 RESULTS A total o f 105 hospital-born neonates v/ho received prophylaxis at birth over a period of 8 months, from November 1995 to July 1996, were included in the study. Ten neonates, bom at home during this period, and who did not receive any prophylaxis, were also included into the study. Initially, a pilot study of 15 cases was carried out. Since the methods used and results obtained were not significantly different from those o f the main study, these 15 cases were included in the analysis of results. The problems relating to the proper transportation of specimens to central SAIMR, identified in the pilot study, were rectified in the main study. Only 95 of the 105 hospital-born neonates were used in the calculation o f the percentage o f cases attributable to chlamydia. This was because in 5 cases, specimens reached the central SAIMR too late for culture and in other 5 cases, chlamydia could not be cultured due to contamination by other organisms. The 10 home-bom neonates were excluded from all the analyses o f aetiological causes o f neonatal conjunctivitis. They were included in the study mainly to compare the organisms causing their conjunctivitis with hospital -born neonates who received prophylaxis. In summary, the incidence percentage o f chlamydia was calculated using 95 cases and that of the other organisms are calculated out of ORGANISMS ISOLATED Table 4.1 shows the bacteria isolated from the 105 hospital- bom neonates, and their day of presentation. /21

32 TABLE 4.1: BACTERIA ISOLATED ON CULTURE OF SWABS FROM HOSPITAL BORN BABIES CASE NO CHLAMYDIA CULTURE OTHER BACTERIAL ISOLATES ON ROUTINE CULTURE DAY OF PRESENTATION WITH CONJUNCTIVITIS AFTERBIRTH Case 1 Positive Staphylococcus epidermidis S* day Case 2 Negative Staphylococcus aureus. Neisseria sonorrhoeae (Beta-lactamase negative) 4th day Case 3 Negative Haemophilus influenzae 30th day Case 4 Negative Staphylococcus aureus Klebsiella oneumoniae Case 5 Negative Staphylococcus epidermidis Neisseria gonorrhoeae (Beta-lactamase negative) Bacillus sp e c i.: 2'" day 13th day Case 6 Specimen unsuitable for culture due to late arrival Streptococcus pneumoniae and Staph epidermidis 3"" day Case 7 Negative Staphylococcus epidermidis Streptococcus pvogenes 11th day Case 8 Negative Staphvlococcus epidermidis 8th day Case 9 Specimen unsuitable for culture due to late arrival Staphylococcus epidermidis 1st day Case 10 Negative Staphvlococcus epidermidis Streptococcus pneumoniae 2nd day Case 11 Specimen unsuitable for culture due to late arrival Staphvlococcus epidermidis 4th day,/22

33 CASE NO CHLAMYDIA CULTURE OTHER BACTERIAL ISOLATES ON ROUTINE CULTURE DAY OF PRESENTATION WITH CONJUNCTIVITIS AFTERBIRTH Case 12 Specimen unsuitable for culture due to late arrival Staphvlococcus aureus 9th day Case 13 Positive Staohvlococcus aureus Streptococcus viridans Case 14 Negative Neisseria gonorrhoeae (Beta-lactamase positive) Case 15 Negative Staphvlococcus epidermidis Streptococcus viridans Case 16 Negative Haemophilus influenzae (Beta-lactamase negative) Case 17 Negative Neisseria gonorrhoeae. (Beta-lactamase negative) Staphvlococcus aureus 15th day 7th day 3"" day 4th day 2 ^ day Case 18 Negative Staphvlococcus aureus K f day Case 19 Negative Staphvlococcus epidermidis 15th day Case 20 Negative N o growth o f any bacteria 9th day Case 21 Negative Staohvlococcus epidermidis Bacillus species Streotococcus viridans Case 22 Negative Neisseria gonorrhoeae. (Beta-lactamase negative) 6th day 2 " 'day Case 23 Negative Staohvlococcus epidermidis 7th day Case 24 Negative Staphvlococcus epidermidis 7th day Case 25 Negative Staphvlococcus aureus 2nd day 723

34 23 CASE N O CHLAMYDIA CULTURE OTHER BACTERIAL ISOLATES ON ROUTINE CULTURE DAY OF PRESENTATION WITH CONJUNCTIVITIS AFTERBIRTH Case 26 Negative Neisseria gonorrhoeae. (Beta-lactamase negative) 4th day Case 27 Negative Klebsiella oneumoniae 2 ^ day Case 28 Negative Staphvlococcus epidermidis 14th day Case 29 Negative Enterobacter species 10th day Case 30 Negative I :aphvlococcus epidermidis 5th day Case 31 Negative Staphvlococcus epidermidis Streptococcus viridans 3 rd day Case 32 Negative Staphvlococcus epidermidis 3^ day Case 33 Positive Staphvlococcus aureus Streptococcus pneumoniae 10th day Case 34 Negative Staphvlococcus epidermidis 12th day Case 35 Negative Staphvlococcus epidermidis 2 ^ day Case 36 Negative Staphvlococcus epidermidis 16th day Case 37 Negative Staphvlococcus epidermidis 7th day Case 38 Positive Staohvlococcus epidermidis 8* day Case 39 Positive Staphvlococcus aureus 25th day Case 40 Negative Staphvlococcus epidermidis Neisseria gonorrhoeae (Beta-lactamase negative) 4^ day Case 4 1 Specimen unsuitable for c ltu re due to late arrival Staohvlococcus aureus 15th day,/24

35 CASE NO CHLAMYDIA CULTURE OTHER BACTERIAL ISOLATES ON ROUTINE CULTURE DAY OF PRESENTATION WITH CONJUNCTIVITIS AFTERBIRTH Case 42 Negative Haemophilus influenzae (Beta-lactamase negative) Case 43 Negative Neisseria sonorrhoeae (Beta-lactamase negative) Staphvlococcus epidermidis Case 44 Negative Haemonhilus influenzae (Beta-lactamase negative) 21st day S* day 2 ^ day Case 45 Negative Staphvlococcus enidermidis 14th day Case 46 Negative Staphvlococcus epidermidis 9th day Case 47 Negative Neisseria sonorrhoeae (Beta-lactamase negative) Staphvlococcus epidermidis Case 48 Negative Staphvlococcus epidermidis Serratia marcescens Case 49 Positive Streptococcus viridans Staphvlococcus epidermidis Case 50 Negative Neisseria sonorrhoeae (Beta-lactamase negative) 3rd day 3rd day 3rd day 3rd day Case 51 Negative Staphvlococcus aureus 8th day Case 52 Negative Staphvlococcus epidermidis 2=" day Case 53 Positive Staphvlococcus epidermidis 14th day Case 54 Negative Staphvlococcus epidermidis 3rd day Case 55 Negative Staphvlococcus enidermidis Bacillus species Case 56 Negative Streptococcus oneumoniae Klebsiella pneumoniae 5th day S"1day,/25

36 25 CASE N O CHLAMYDIA CULTURE OTHER BACTERIAL ISOLATES ON ROUTINE CULTURE DAY OF PRESENTATION WITH CONJUNCTIVITIS AFTER B R T H Case 57 Negative Staphvlococcus epidermidis Haemophilus influenzae (Beta-lactamase negative) Case 58 Negative Neisseria sonorrhoeae (Beta-lactamase positive) Staphvlococcus epidermidis Bacillus species Case 59 Negative Staphvlococcus epidermidis Proteus mirabilis 14* day 11* day 2 ^ day Case 60 Negative Klebsiella pneumoniae 2nd day Case 61 Negative Staphvlococcus epidermidis Bacillus species 15* day Case 62 Negative Staphvlococcus epidermidis 5* day Case 63 Could not be isolated due to contamination Staphvlococcus epidermidis 8* day Case 64 Could not be isolated due to contamination K lebsiula pneumoniae Proteus mirabilis Staphvlococcus epidermidis 2nd day Case 65 Negative Staphvlococcus epidermidis 2nd day Case 66 Could not be isolated due to contamination Staohvlococcus enidermidis 18* day Case 67 Negative Bacillus species Staohvlococcus enidermidis Case 68 Negative Bacillus snecies Staohvlococcus enidermidis 14* day 2"" day./26

37 26 CASE NO CHLAMYDIA CULTURE OTHER BACTERIAL ISOLATES ON ROUTINE CULTURE DAY OF PRESENTATION W ITH CONJUNCTIVITIS AFTERBIRTH Case 69 Negative Streptococcus pneumoniae Staphvlococcus epidermidis 4th day Case 70 Negative Staphvlococcus aureus 3rd day Case 71 Negative Streptococcus viridans Staphvlococcus epidermidis 13th day Case 72 Negative Staphvlococcus aureus 4th day. Case 73 Negative Staphvlococcus epidermidis 2 ^ day Case 74. Negative Staphvlococcus epidermidis 2 ^ day Case 75 Negative Staphvlococcus aureus 2 ^ day Case 76 Negative Staphvlococcus aureus Bacillus species 5th day Case 77 Negative Staphvlococcus epidermidis 7* day Case 78 Negative Staphvlococcus epidermidis nd day Case 79 Negative Neisseria sonorrhoeae (Beta-lactamase negative) Bacillus species Staphvlococcus enidermidis Case 80 Negative Staphvlococcus aureus Staphvlococcus epidermidis 5th day 8th day Case 81 Negative Streptococcus pneumoniae 4th day Case 82 Negative Staphvlococcus epidermidis Proteus mirabilis 7lh day Case 83 Negative Staphvlococcus aureus 20th day Case 84 Negative Pseudomonas aerueinosa Staphvlococcus enidermidis 4th day -

38 CASE N O CHLAMYDIA CULTURE OTHER BACTERIAL ISOLATES ON ROUTINE CULTURE DAY OF PRESENTATION WITH CONJUNCTIVITIS AFTERBIRTH Case 85 Negative Streptococcus viridans Staphvlococcus aureus Case 86 Positive Staphvlococcus aureus Staphvlococcus epidermidis 3"" day Not known Case 87 Positive Staphvlococcus epidermidis 2 ^ day Case 88 Negative Pseudomonas aerueinosa Staphvlococcus epidermidis 9th day Case 89 Could not be isolated due to contamination Klebsiella pneumoniae Escherichia coli Z ^day Case 90 Positive Klebsiella pneumoniae Staohvlococcus aureus 8th day Case 91 Negative No growth o f any bacteria 5th day Case 92 Negative Staphvlococcus aureus 1st day Case 93 Negative Proteus mirabilis Staphvlococcus aureus 1st day Case 94 Negative No growth o f any bacteria 1st day Case 95 Positive Staphvlococcus aureus 12th day Case 96 Negative Staphvlococcus aureus Streptococcus pvoaenes 6th day Case 97 Negative Staohvlococcus enidermidis 3rd day Case 98 Negative Staohvlococcus epidermidis 16th day Case 99 Negative Neisseria sonorrhoeae CBeta-lactamase negative) Bacillus species Staohvlococcus aureus 10'h day 728

39 28 CASE NO CHLAMYDIA CULTURE OTHER BACTERIAL ISOLATES ON ROUTINE CULTURE DAY OF PRESENTATION WITH CONJUNCTIVITIS AFTERBIRTH Case 100 Positive Staphvlococcus aureus 3rd day Case 101 Negative Staohvlococcus epidermidis Streptococcus oneumoniae 2 ^ day Case 102 Negative Streptococcus pneumoniae 10th day Case 103 Negative Escherichia coli Staphvlococcus epidermidis Case 104 Negative Staphvlococcus epidermidis Streptococcus oneumoniae Citrobacter freundii 4til day 5th day Case 105 Could not be isolated due to contamination Streotococcus pneumoniae Staohvlococcus epidermidis 5th day A variety of organisms were isolated and in 19 of 105 patients (18%) multiple pathogens were identified. A summary o f the organisms isolated is given in table 4.2. S. epidermidis was the most common organism identified (57%). S. aureus was cultured in 25 cases (23,8%) and was the most important pathogenic organism isolated.. I n 7 of 25 cases (28%) this was the only organism isolated. S. aureus also co-existed in 7 of the 12 chlamydia, positive cases. O f 25 S. aureus isolates, 24 were sensitive to erythromycin and chloramphenicol. No growth o f any bacreria was recorded in 3 cases where the possibility exists that a viral agent may have played a role in the aetiology o f conjunctivitis. 129

40 29 TABLE 4.2: SUMMARY OF ORGANISMS ISOLATED FROM EYE SWABS OF HOSPITAL BORN NEONATES fn - 105) ORGANISM NUMBER PERCENTAGE S. epidermidis S. aureus *C. trachomatis 12 o f N. sonorrhoeae S. pneumoniae S. viridans K. pneumoniae H. influenzae P. aerueinosa Enterobacter species E. coli S. pvoaenes P. mirabilis S. marcescens C. freundii Bacillus spp No growth o f any bacteria Ten cases excluded while calculating percentage for Chlamydia due to delayed transportation (5 cases) and contamination of cell cultures (5 cases) 730

41 N. gonorrhoeae was isolated in 12 cases. In 2 o f the 12 (16%) the organisms were Beta-lactamase-positive, penicillinase producing H gonorrhoeae (PPNG). All 12 isolates were susceptible to a combination of erythromycin and chloramphenicol. In 4 cases it co-existed with Bacillus spp and in 6 cases with Sh epidermidis. C. trachomatis was isolated in 12 of the 95 cases (12.4%). In the pilot study, a delay in transport prevented successful culture in 5 cases, and in 5 cases in the main study contaminmion by other organisms prevented the growth o f chlamydia. Thus, 10 specimens were excluded from the analysis o f chlamydia. Seven o f the 12 patients presented during the 2nd week o f life. The other 5, who had mixed infections, presented earlier. S. epidermidis was isolated in 57% of cases. This organism is usually considered to be a contaminant. In the majority o f cases it co-existed with other relevant and potentially pathogenic organisms. In 25 cases it was the only organism isolated. It is arguable whether S. epidermidis played any role in causing the conjunctivitis in these patients. Streptococcus pneumoniae was the fourth most important pathogen and was detected in 10 (9.5%) o f cases. The time o f presentation varied from 2 days to 15 days. Streptococcus viridans were also identified in 6.6% o f cases. Several Gram-negative bacilli, the most important of which was Klebsiella pneumoniae (6.6%) were isolated. Isolation rates of E. coli. Enterobacter spp and Pseudomonas aeruginosa were low - less than 3% each. However, in total these organisms accounted for 16% o f positive cultures.... /3 1

42 Haemophilus influenzae was found in 4.7% of cases. Haemophilus aeavptius was not isolated in this study. S. pvogenes was isolated in 1.9% o f cases. 4.2 TIME OF PRESENTATION The time o f presentation in the hospital born infants is shown in Figure I. Only 104 cases were included in the analysis because one mother could not remember the day of onset o f conjunctivitis. It shows that the largest number of cases presented on the second day o f life; 72% of cases presented in the first week, 25% in the second and only 3% thereafter. Among the important pathogenic bacterial conjunctivitis presenting after first week, there were 12 cases o f S. aureus. 8 cases of C. trachomatis. 3 cases of N. gonorrhoeae. 2 cases o f Sh pneumoniae. 2 cases of ft, viridans. 1 case of K. pneumonia. 3 cases o f H, influenzae. 1 case each of P, aerogenosa. ft pvogenes and Enterococci. Figure II focuses on seven important bacteria and their week of presentation. It is clear that all organisms most frequently caused infection in the first week after birth except for chlamydia which presented more often in the second week o f life../32

43 Number of cases N> O Day after birth % o -g to 11O ZJ 0) ZJ (D 0 Z3 1 O o 5. c o CZ) I -1 CF # Fig 1: Day of presentation of - - / 33

44 -33- v/////////////////m^ ^ 5 a^-a av-iraf.; Bx: ssksssatff No. of cases X 4 ^ CD CO O Fig 2: Week of presentation of conjunctivitis in hospital born neonates in relation to important organisms isolated (n= 104)

45 ANTIBIOTIC SUSCEPTABILITY The antibiotic sensitivity patterns o f the five most commonly isolated significant pathogenic organisms is shown in table 4.3. Not all the strains were tested against all the antibiotics. Only 4 of the 12 N. gonorrhoeae strains were tested against tetracycline. None o f the chlamydia strains were tested for susceptibility to antibiotics by the SAJMR. TABLE 4.3: SENSITIVITY RESULTS OF 5 MAJOR ORGANISMS ISOLATED IN THE STUDY TO VARIOUS ANTIBIOTICS Organism Number Erythromycin Chloramphenicol Penicillin Tetracy Gentamycin n (%) n (%) cline N (%) n (%) n (%) S. aureus (96) 24 (96) 5 (20) *NT 18 (72) S. pneumoniae (100) 10 (100) 6 (60) 8 (80) NT S. Viridans 7 7 (100) 6 (86) 7 (100) 1 (14) 4 (57) K. pneumoniae 7 1 (14) NT 0 (0) 4 (57) 6 (86) N. gonorrhoeae 12 N T 12 (100) 10(83) 4/4** (100) NT Overall 61 42/49 (85) 52/54 (96) 28 (45) 17/28 (61) 28/39 (71) * N T - Not tested ** Only 4 specimens tested 735

46 INCIDENCE The average number o f deliveries at Sebokeng Hospital during 1996 was approximately 480 per month. The actual number o f hospital deliveries during the eight-month study period was A total o f 105 cases o f conjunctivitis were noted among hospital-born neonates. An additional 36 cases were noted among hospital-born neonates during the period o f the study which had to be excluded from the study for reasons stated earlier, eg. presentation at times outside the study enrolment period. Thus the total number o f hospital-born neonates suffering from conjunctivitis was 141. Thus, the true incidence o f conjunctivitis in the first month of life among hospital-born neonates at Sebokeng Hospital can be calculated as follows: 141 cases per births = 3.66% o f all births or 1 case per 27 births. The chlamydial infection rate was 0.45 cases/100 live births and that o f gonococci 0.41/100 births. 4.5 HOME BIRTHS Ten babies who were born at home, and received no prophylaxis at birth, were enrolled into the study.,/36

47 A comparison was made between hospital and home-deliveied babies regarding the pathogens responsible for eye infections. Organisms isolated from home-born babies are highlighted in table 4.4. R gonorrhoeae was cultured in three cases (30%) among the home delivery group. S. aureus was also isolated in another 3 cases (i.e. 30%). TABLE 4.4: BACTERIA ISOLATED ON CULTURE OF SWABS FROM HOME BORN BABIES CASE. DAY OF PRESENTATION. BACTERIA ISOLATED. CHLAMYDIA Case 1 12th N. gonorrhoeae. S. epidermidis. Bacillus species Specimen arrived late Case 2 gtii H. influenzae. S. epidermidis. Bacillus species Negative Case 3 2nd H. influenzae Negative Case 4 21st S. pneumoniae Negative Case 5 2nd S. aureus Negative Case 6 5th N. gonorrhoeae. S. epidermidis. Bacillus species Negative Case 7 1st N. gonorrhoeae. S. epidermidis Could not be isolated due to contamination Case 8 18th Enterobacter species Negative Case 9 21st S. aureus Negative Case 10 3rd S. aureus. Enterobacter species Negative 737

48 O f the total number o f h( me born infants 50% presented in the first week and 70% in the first two weeks. There appeared to be no difference between the time of presentation of home and hospital-bom infants. 4.6 MANAGEMENT OF CONJUNCTIVITIS All babies enrolled into the study were treated at the time o f presentation, and prior to the culture and sensitivity results being available, with erythromycin syrup mg/kg four times/day for ten days and chloramphenicol eye drops (five drops four times daily) for seven days. In 15 case ( '%) this treatment failed and babies returned with persistently discharging eyes. The causative pathogens isolated in these 15 cases, at the time of first presentation, are shown in table 4.5. Treatment with a single injection of 50 mg/kg Ceftriaxone was given to all these infants. TABLE 4.5: ORGANISMS ISOLATED FROM NEONATES NON-RESPONSIVE TO INITIAL THERAPY ORGANISM NO K. pneumoniae 7 P. aerueinosa 2 E. coli 2 S. aureus 1 N. eonorrhoeae 1 Enterobacter species _2 TOTAL

49 QUESTIONNAIRE RESULTS The questionnaire analyses showed the following data. O f the 1 mothers interviewed : 1. Only 10 banes (9%) were bom at home and none o f them received any eye prophylaxis mothers (22%) were married mothers (78%) complained about vaginal discharge to the clinic sister mothers (10%) who attended the clinic did not complain to the clinic sister about their vaginal discharge. 5. Only 3 mothers (3%) did not attend antenatal clinic mothers (30%) gave a history of receiv 'i..h some form of treatment from the clinic for their vaginal discharge. V39